TL;DR Summary:
There’s a sentence sitting in paragraph two of almost every message I get about this, and it’s the answer to the question being asked in paragraph five.
You’ve been treating the boundary like a lever. It was never a lever. It was an instrument, and it already gave you a reading.
The reason you can’t stop rewording it has a name in learning theory, and it’s the same reason slot machines work. It is not weakness and it is not stupidity.
There’s a variable in the research on why people stay that has almost nothing to do with him, and it’s the one quietly running your entire decision.
One study found it wasn’t the years that made people stay in an unhappy relationship. It was something else you’ve been doing more of every single week.
Being a project still has hope in it. A decision doesn’t. That’s the whole blog, and you’re gonna hate the last layer.
The Message I’ve Read Four Hundred Times
It shows up in my DMs, in my She Rises Community, in comments, in first sessions with clients.
Four paragraphs, sometimes six. A timeline. Occasionally a screenshot.
She’s tried saying it soft. She’s tried saying it direct. She waited for a calm moment because last time she brought it up during a fight and figured that was the problem. She used an “I feel” statement. She used the one from the podcast. She’s read two books on this. She’s in therapy.
And NOTHING is working; she’s tried EVERYTHING…
And then the last line is always the same: What should I do? Or should I give him a consequence?
Here’s the part that took me a few hundred of these to notice.
The answer is already in the message. Usually by paragraph two. Written by her, in her own words, in more detail than I could ever collect myself.
She has documented eighteen months of behavior with better recall than most people have for their own childhoods.
She knows what he said in March. She knows what he did in June. She knows exactly how long the good stretch lasted after the last big conversation.
You’ve tried EVERYTHING except the ONE THING that will change anything.
So this isn’t a blog about what to say to him. There are nine thousand of those, and you’ve read most of ‘em.
This is about what you’re doing with answers you already have, and why you can’t act on them yet.
We’re gonna go one layer at a time, step by excruciating step.
Let’s dive in!
Your Experiments Are Broken
You came for a better sentence. Or a better boundary, or a consequence framework, or the magic phrasing that finally gets through.
And I think some part of you is expecting me to say your tools weren’t good enough.
Nah.
The tools worked perfectly. That’s the problem. Every single boundary you set came back with a clean result, and you didn’t like the result, so you ran it again.
A boundary is an experiment.
You set it, and then you watch what happens, and what happens tells you who you’re dealing with.
It’s a diagnostic that you’ve been swinging like a crowbar.
Which means you’re not failing to communicate, and you’re not bad at this.
You’re running broken experiments. Dozens of them, actually.
They’re broken in four specific ways, and underneath each is a reason you need it broken.
Layer One: The Bad Experiment
Let’s start with the mechanics, because they’re the least threatening part and they’re also embarrassingly easy to verify against your own text history.
Here’s what a bad experiment looks like:
You change every variable at once. New phrasing, new tone, new timing, plus you’ve also decided to be less reactive this month. Something shifts a little, and you have no idea which input did it. Neither does he.
You never define failure in advance. You never once decided, before the conversation, what result would mean “this isn’t happening.” So no result can ever mean that.
There’s no endpoint. No round count, no window, no deadline. An experiment with no stopping rule cannot produce a conclusion. It can only produce another round.
A bad result triggers a redesign instead of a decision. Nothing changed after the last talk, so you refine the talk. That’s not analysis, that’s a rewrite.
Notice that all four of those keep the experiment open.
An experiment with no falsifiable outcome isn’t an experiment.
It’s a ritual. And rituals aren’t designed to produce findings.
They’re designed to make you feel less powerless while you wait…
Layer Two: You’re Not Missing the Data. You’re Disqualifying It.
Okay, so why doesn’t the bad design get fixed?
Because you’re not actually missing the results.
You saw every one of them in real time. And then you threw each one out with a reason.
Nobody comes into a first session and says “he doesn’t care that this hurts me.” Nobody.
In thousands of sessions I’ve heard it maybe twice, and both times the woman was already halfway out the door.
What I hear instead:
“He’s under a lot of stress right now.”
“He never had this modeled for him growing up.”
“I probably came in a little hot, so.”
“He’s just not great with emotional stuff.”
“He did do it that one time, though.”
“I don’t think he even realizes he’s doing it.”
Every one of those might be true. That’s what makes them work.
This is motivated reasoning, and it’s one of the most replicated findings in social psychology.
You don’t reason toward the truth. You reason toward the conclusion you can afford, and then you stop looking.
The searching feels exactly like thinking. It feels rigorous. It feels fair-minded, even.
And notice the direction of the excuses.
Every single one moves the cause off of “what he chose” and onto something neither of you can be blamed for.
Stress. Upbringing. Her tone.
This is not a perception problem. Your perception is excellent.
It’s just aimed at a conclusion you picked before you started looking…
Layer Three: It Worked Once
Now we get to the compassionate part, because the disqualifying isn’t delusional.
It’s evidence-based, which is what makes it so sticky.
He did change once for six days, or two weeks, or that whole month in the spring.
And it was real. He wasn’t running a con. He meant it when he said it, and he did it, and it was so good that you can still describe that stretch in detail.
Then it faded. And every conversation since has been an attempt to get back to that month.
In learning theory, the most persistent behavior on earth isn’t the behavior that gets rewarded every time.
It’s the behavior that gets rewarded unpredictably.
Rats, pigeons, humans, slot machines… Same finding since the 1930s: when the payoff is intermittent, the behavior takes dramatically longer to extinguish.
The not-knowing is the engine. Predictable failure is easy to quit. It’s the maybe that keeps you pulling.
To be clear, this is lab work on reinforcement schedules, so I’m applying it to your relationship by analogy, not claiming somebody ran this study on your boyfriend, but you get my point.
Here’s where most people get this wrong and start writing him off as a manipulator.
He probably wasn’t performing. Those six days weren’t bait.
They were the top of his capacity, and capacity is not a promise. It’s a ceiling.
He showed you what he can do at maximum effort under maximum motivation, and it turns out he can’t live there.
That’s not evil… That’s a nervous system with a limit.
But you filed it under “proof of what’s possible,” when the actual finding was “this is the sample he cannot sustain.”
Same six days. Two completely different readings.
One of them keeps you here for another year.
Layer Four: You’d Rather Be a Project
Why is that, you ask? Easy.
Rewriting the sentence is the activity that lets you avoid the decision.
And you’d rather be a project than a person who has to grieve, because a project still has hope in it and a decision doesn’t.
As long as you are the variable, the outcome is still open.
As long as the outcome is open, nothing has to be mourned.
There’s still a version where you find the words, and he softens and the last three years turn out to have been the hard part of a good story.
Making yourself the project isn’t laziness. It’s the opposite, and that’s exactly why it’s so effective.
It’s effortful, sincere, expensive, and completely uncriticizable.
Nobody in your life is gonna tell you to stop working on yourself.
Your therapist won’t. Your friends won’t.
I, on the other hand, have no issue telling you this :)
It’s the most sophisticated stalling mechanism available to a high-achieving woman, because it’s indistinguishable from growth.
In fact, when a research team tested this with about nine hundred people, giving them an unhappy relationship scenario and varying what had already been invested, time invested didn’t reliably move the decision to stay.
Effort did.
It’s not the years that trap you. It’s the work.
Every draft you write raises the cost of concluding it didn’t work, which makes you more likely to write another one.
The loop feeds itself.
You are not stuck because you haven’t tried hard enough. You’re stuck partly because of how hard you’ve tried.
Me Too…
I did this in my last relationship…
I worked on myself and the relationship, like it was my job.
I got more emotionally available. I got better at regulating instead of going quiet. I worked on being less passive, on asking for what I wanted, on all the stuff I teach.
And I was genuinely improving. That part wasn’t fake.
What I was not doing was looking directly at the fact that her capacity was shrinking. Month over month. Less presence, less reaching, less of her in the room.
I had the data. I could’ve drawn you the graph.
But I had a much more comfortable project available, which was me.
And as long as I was the variable, I never had to sit down and ask what I was gonna do about a partner who had less for me every month.
Then it ended, “out of nowhere.”
Except it wasn’t out of nowhere.
It was out of eleven months of somewhere that I had watched happen and filed under “things I can fix by being better.”
Blindsided is what I called it for a long time.
What actually happened is that I ignored the information that was right in front of my face because accepting it would mean I’d have to make a very difficult decision.
Layer Five: You Already Thought You Were the Problem
This is the floor. Everything above has been sitting on this the whole time.
Making yourself the project only feels natural if some Part of you was already pretty sure you were the reason.
Yep, I said it.
And that Part did not show up when he did. She was there WAY before him.
Nobody says this out loud, obviously. In a first session it comes out like:
“I know I’m a lot.”
“I just need to work on my communication.”
“I don’t wanna be needy about it.”
“I’m honestly not the easiest person to be with.”
“He’s got his stuff, but I’ve got mine too, so.”
Underneath all five of those is one belief, and it’s older than this relationship: love is something you earn by being enough, and when it goes wrong, the math says you weren’t.
Which is why the project feels safe.
It doesn’t threaten the belief. It confirms it.
Every draft you write is one more piece of evidence that yes, this is on you, and yes, you can fix it if you’re good enough, and that hurts a lot less than the alternative.
The alternative being: he’s not giving you what you asked for, you are not the reason, and you’d have to leave a relationship you can’t prove you did anything wrong in.
And me? Same floor, different room.
What I actually believed, underneath eleven months of self-improvement, was that her capacity was a referendum on me.
That if I got emotionally available enough, regulated enough, present enough, she’d have more to give.
Her shrinking was a scoreboard, and I was losing, and I could still win.
It never once occurred to me that she just had less, and it wasn’t about me at all.
You cannot work your way out of a worthiness problem.
That’s why nothing you said worked, and it’s why nothing you say next will either.
The sentence was never the variable.
What you believe you’re allowed to want is the variable.
Ok, Cody, thank you for dissecting my brain yet again…
What can I DO about this???
Great question, so glad you asked :)
The Evidence Protocol
Five steps. Four of them are the things you’ve been skipping, in order.
The fifth is the one that makes the other four possible, and it’s the only one that isn’t about him at all.
Oh, and none of them involves finding better words.
Step 1: Listen
Not to what he says. To what has happened since the last time you said it.
Pick a window. Thirty days, backwards from today.
Write down what actually occurred, behavior only, no interpretation and no context you added to make it make sense.
He said he’d handle it. Did the thing happen or not?
Written down matters. Your memory is not neutral here, and it will smooth this for you if you let it work from vibes.
Step 2: Discern
Now sort what you wrote into two piles: can’t and won’t.
Can’t is a capacity problem. He doesn’t have the skill, the bandwidth, the nervous system regulation.
It’s real, it deserves compassion, and it is still not your project to solve on his behalf.
Won’t is a willingness problem. He understood, he was able, and he chose the version that cost him less.
Only one of these has a path forward, and mistaking one for the other is why you’re on round nine.
You’ve been treating a won’t like a can’t for a very long time, because can’t is the kinder read and you are a kind person.
Step 3: Accept
The one everybody skips. And by everybody I mean you specifically. 😅
Accepting isn’t approving, and it isn’t forgiving, and it definitely isn’t deciding you were wrong to want what you wanted.
It’s ending the practice of treating the last eight months as a data collection error.
The results are the results.
They were the results in March too.
Step 4: Act
Sometimes that’s leaving; oftentimes, unfortunately, it is.
Sometimes it’s a consequence that you genuinely intend to follow through on, which is different from a consequence you announce and then quietly withdraw when he gets sad.
If you don’t intend to do it, don’t say it, because now you’ve taught him what your words are worth.
But it’s for sure not running another broken experiment; it’s time to LISTEN to the data.
Step 5: Go Get the Part That Set the Number
Here’s the problem with everything I just gave you.
You can run the first four steps perfectly and still not move.
And then you’ll stay anyway, because Layer Five hasn’t budged an inch.
So this last one goes inward, and it’s slower than the rest, and it’s the only step here that actually changes what you’re able to do with the evidence.
Somewhere in you is a part that decided a long time ago that love has to be earned.
She’s usually young. She’s not stupid, and she’s not dramatic.
She built that conclusion out of real evidence, in a house or a friendship or a first relationship where the evidence honestly did point that way.
She was right, back then, about the room she was standing in.
She’s just never been told the room changed.
You can’t argue her out of it. People have been trying for years.
Your friends have made the case, your therapist has made the case, I’m making it right now in a blog, and none of us have moved her a millimeter, which should tell you something about the method.
So don’t argue.
Do this instead (Yes, this is IFS):
Find her in your body. Imagine walking away for good, done, no more drafts. Notice where that lands physically. Chest, throat, stomach, the back of your neck. Don’t analyze it, just locate it. That’s her.
Check how you feel toward her. Annoyed? Embarrassed? Wanna tell her to shut up and get it together? See if you can extend some Curiosity and Compassion toward her instead.
Unblend. “This is a part of me. Not all of me.” Out loud, if you’re by yourself. Yes, it feels ridiculous. Do it anyway.
Ask, don’t fix. How old is she? What does she think happens if she’s wrong about what you’re worth? Then shut up long enough to actually hear an answer, which for a lot of you is gonna be the hardest instruction in this entire blog.
Update her. She’s been running on information collected from people who are not in your life anymore. She doesn’t know what you know now. She’s never once been shown the evidence you’ve spent the last year collecting. Bring her up to speed.
Steps one through four tell you what the evidence says.
This one decides whether you’re allowed to act on it.
These are small, unglamorous reps and none of them will feel like a breakthrough.
But they are how you stop being a person who collects evidence and start being a person who uses it.
You Got Your Answer a Long Time Ago
You came in asking for a better way to say it.
What you’ve actually been doing is requesting a second opinion on a result that came back clearly the first time, and then a third, and then a ninth.
The evidence was never ambiguous. Your willingness to look at it and act on it was.
And I’m not gonna tell you what to do with it, because I don’t know your life and because the whole point is that the reading is yours.
You collected it. You’re the only one who was in the room for all of it.
Nobody else gets to deliver this verdict, including me, and that’s the terrible part and the good part at the same time.
You don’t need a better way to say it.
You need to believe the answer you already got.
You’ve got this, and until next time… Live Heroically 🧠
Ready to Complete Heartbreak in 63 Days?
Start Here: She Rises Community.
Inside is a 63-day program of the exact things I did myself to go from wrecked to regulated using neuroscience.
Plus, You’ll Have Access To:
A 63-day program to go from wrecked to regulated
Live coaching calls with me every week
A team of IFS practitioners
Private workshops and masterclasses with therapists and scientists
1-on-1 sessions
A room full of women doing this beside you
If you’re ready to become emotionally available, develop secure attachment, and attract healthy potential partners, this is for you!
The first 7 days are free. Join below!
Supporting Research
Le, B., & Agnew, C. R. (2003). Commitment and its theorized determinants: A meta-analysis of the Investment Model. Personal Relationships, 10(1), 37-57.
Rusbult, C. E., Martz, J. M., & Agnew, C. R. (1998). The Investment Model Scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5(4), 357-387.
Tran, P., Judge, M., & Kashima, Y. (2019). Commitment in relationships: An updated meta-analysis of the Investment Model. Personal Relationships, 26(1), 158-180.
Goodfriend, W., & Agnew, C. R. (2008). Sunken costs and desired plans: Examining different types of investments in close relationships. Personality and Social Psychology Bulletin, 34(12), 1639-1652.
Rego, S., Arantes, J., & Magalhães, P. (2018). Is there a sunk cost effect in committed relationships? Current Psychology, 37(3), 508-519.
Arkes, H. R., & Blumer, C. (1985). The psychology of sunk cost. Organizational Behavior and Human Decision Processes, 35(1), 124-140.
Staw, B. M., & Fox, F. V. (1977). Escalation: The determinants of commitment to a chosen course of action. Human Relations, 30(5), 431-450.
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480-498.
Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2(2), 175-220.
Ferster, C. B., & Skinner, B. F. (1957). Schedules of reinforcement. Appleton-Century-Crofts.
Humphreys, L. G. (1939). The effect of random alternation of reinforcement on the acquisition and extinction of conditioned eyelid reactions. Journal of Experimental Psychology, 25(2), 141-158.
Schwartz, R. C., & Sweezy, M. (2020). Internal Family Systems therapy (2nd ed.). Guilford Press.
Leave a Comment
What data have you been ignoring? I know it just popped into your head… Tell me below!
This article is educational in nature and not a substitute for therapy. If attachment wounds or relational trauma are impacting your well-being, working with a trauma therapist can help your nervous system relearn safety in connection.